The 2018 Voting System Referendum in British Columbia: Suggestions for a Fair Process

A Brief for the Attorney General of British Columbia

Prepared by Dr. Dennis Pilon, Associate Professor, Department of Politics, York University

February 2018

Introduction

The government has chosen to use a referendum to have the public decide whether BC should adopt a new voting system. While this approach can lend significant legitimacy to the outcome, referendums are often blunt instruments for making public decisions, subject to a host of challenges similar to those at work in any situation where the public is voting (i.e. a general election) but also different in important ways. It is incumbent upon the government to assure that the process is fair, equitable, and seen to be legitimate. Fairness refers to the openness of the outcome – that no choice is seen to be privileged in the process. Equity refers to the openness of the process in terms of garnering broad participation. Legitimacy refers to gaining a significant degree of public support for the process and outcome. In addition, the outcome itself may pose potential process-oriented challenges for the government if the public votes to support a change in the voting system, specifically as concerns how best to facilitate the design and implementation of a new system.

To that end, this brief will address two broad areas of concern: referendum issues and voting system design and implementation issues. It will draw on academic research related to these questions as well as the concrete experience of referendum processes and voting system use across western countries. In other words, it will attempt to draw from fact-based, systematic research on, and real world experience with, referenda and voting systems to frame how we should approach questions of referendum process, ballot structure/design, voter turnout, public information and education, and voting system design and implementation.

1. Referendum issues

Referendums allow governments to delegate responsibility for making decisions directly to the public, thus increasing the potential legitimacy of whatever decision results. But the process by which a referendum is carried out may crucially affect whether the outcome is, or is seen to be, legitimate. There are three broad areas of concern here that the government needs to address involving questions of ballot design, public participation, and public education. a. Ballot design

The wording of the question and the structure of choices on the ballot are crucial decisions in assuring that the outcome really does reveal what the voters prefer in terms of making a choice about the provincial voting system. There are at least five possible ballot designs that could be employed: (1) a choice between the status quo and a specific alternative; (2) a choice between the status quo and a generic alternative; (3) a ranked choice amongst the status quo and various alternatives; (4) a choice between the status quo and various value alternatives; (5) a choice between the status quo and a generic alternative, followed by a nominal or ranked choice amongst specific alternatives.

1 Opponents of change have called for option 1, which does appear to have the advantage of clarity: the decision would confirm public support for the status quo or for a specific alternative voting system. But as there a number of possible alternative voting systems to the status quo, option 1 runs the risk of dividing those who favour change on the basis of the single option offered as an alternative. In other words, as a ballot question, option 1 might not effectively reveal what the voters really want (i.e. most voters might not want to keep the status quo but not agree on the proposed alternative). Option 2 gets around this by posing the status quo against a generic alternative e.g. a proportional voting system. But critics argue that option 2 would suffer from insufficient legitimacy in terms of the outcome as the public would not have much input on the specific voting system that would be chosen. Such criticism is weak for reasons that will be explored in greater detail below (mostly related to the public’s low level of knowledge of about electoral system details generally) but nevertheless the government must be mindful of how option 2 might be contested and politicized by those opposed to any change. Option 3 is the approach that was utilized in Prince Edward Island’s provincial referendum on the voting system in 2016. Voters were offered a range of five voting systems to choose from, which included the province’s status quo single member first past the post voting system and four alternatives. The referendum used the ‘alternative vote’ form of ranked balloting to allow voters to indicate a hierarchy of preferences over the various choices. Then, through a process of elimination, one option emerged as the choice of a majority of voters. While the experience in PEI worked well, it runs the risk of collapsing two decisions (i.e. ‘do you want change?’ and ‘what kind of change do you want?’) into one in a way that may leave the public will unclear.

Options 1 through 3 could be all used for BC’s referendum on the voting system but they are not ideal choices. Option 1 may fail to reveal what the public really wants and, as such, may lack legitimacy for those interested in change. Options 2, while arguably in line with what research tells us about what voters do or don’t know about election-related institutions, would be open to challenge by those opposed to change who would object to government making all the decisions about the specific alternative voting system to be used, if the vote supported change. Option 3, depending on the choices on the ballot, could produce an outcome where the public’s desire for change is not effectively represented in the outcome. As such, only options 4 and 5 are realistic choices from both a public legitimacy and political contestability standpoint.

Option 4 adds some specificity to the alternatives available to be chosen in terms of the values that would animate the choice e.g. maintaining single member constituencies as a key component, or allowing voters to rank their choices in a multi-member constituency, etc. Voters would not choose the specific system but government would be constrained in making the choice of an alternative voting system by the value priorities established by the voters. Research supports the view that voters can get up to speed on the broad trade offs represented by different voting systems, even if they cannot answer questions about how such systems operate to produce their results. Option 5 is the approach that was utilized in New Zealand in their national indicative referendum on the voting system in 1992. Voters were first asked whether they wished to retain the status quo single member first-past-the-post voting system or opt to change the voting system. Then they were offered a range of four alternative voting

2 systems to choose from. Each of these options could work for choosing BC’s voting system and, in the case of option 5, has been demonstrated to actually work in a referendum.

There are some challenges in taking up either option 4 or 5, some of which are legitimate while others are not. On the ‘not legitimate’ side, various commentators in BC’s media in the fall of 2017 argued that utilizing ballot structures like options 4 would mean that the public is not directly choosing the new voting system. While technically correct, option 4 does balance the decision-making responsibility in a way that matches what research shows to be an affective trade off between public knowledge and expert knowledge. Critics have also argued that approaches like option 5 would be unfair as it would bias the outcome in favour of change. But such claims were largely asserted, rather than persuasively argued. For instance, none made clear just why a two-part question with or without a ranked ballot structure would bias the outcome. Looking to past practice, if anything, historic uses of ballot structures similar to option 5 for voting system referendums in Vancouver in the 1970s and 1990s were seen by their critics as privileging the status quo, not change. Nor were there complaints that option 5 was biased toward change in the experience with it in New Zealand. Other critics have argued that an option 5 approach would not produce a clear outcome on the second question. Indeed, they argue that all attempts to secure a majority decision by using a ranked ballot produce illegitimate results. Needless to say, such views are hard to sustain given the widespread use of majority voting worldwide and in Canada within most political organizations (for leadership and candidate selection, for instance).

The real challenge in taking up options 4 or 5 is simply the documented reality that voters know little about the details of our electoral institutions like voting systems and that it is unlikely, regardless of whatever public education strategy is taken up, that that will change before the fall of 2018 in BC. What this means is that voters will face a challenge in navigating the choices on a ballot like options 4 or 5. However, despite this, there are trade-offs that still make choosing option 4 or 5 the best one for the government. The first is that even though the public is unlikely to become very knowledgeable about the specifics of different voting systems, evidence suggests they can get up to speed on the broad trade-offs involved in the use of different systems, and they can use proxies – sources they trust, like certain media, political parties, or pundits – to help them make their choices. The evidence from the New Zealand’s experience suggests that voters can navigate such ballots and that the results do broadly correspond with what voters actually want to happen. Option 4 would be contested by those opposed to change because voters would not be choosing the specific system, but government could defend the choice as a pragmatic compromise for reasons sketched out above. On the other hand, option 5 would see voters indicate the specific voting system they prefer, which would cast more legitimacy on the outcome and forestall criticisms that change is not voter- driven. Between the two, option 5 is really the gold standard for process but option 4 could also be made to work.

Recommendation: that the government utilize a ballot structure similar to option 5. Barring that option 4 would be the next best choice.

3 b. Public participation

Broad public participation in the referendum will be important for legitimizing the outcome. Low voter turnout, or uneven voter turnout in terms of demography or partisan orientation, will cast doubt on whether the outcome is a legitimate representation of what British Columbians want. Academic research on voter participation in general elections and referendums has revealed some consistent tendencies in terms of who votes and why, as well as some best practices to address these challenges. It should be noted at the outset that the decision by the government to hold the referendum as a stand alone event with a mail-in ballot only compounds the challenges of maximizing and equalizing voter turnout, for reasons that will be set out below. Still, there are decisions that the government can take to increase the likelihood of more equitable participation in the referendum.

Some public participation challenges are general issues related to any electoral event while others are specific to referenda or mail-in balloting. First, it must be stated that media characterizations of individual voters carefully researching political issues and political party manifestos to then make an informed choice on election day does not square with academic research on how actual voters interact with electoral events. As a rule, voters are not well informed on details of policy. Instead, voters widely use information short-cuts – direction from media or political parties they support – to align their general ideological values with specific partisan choices. But whether they actually participate in an electoral event is also influenced by a host of other factors. There are some voters who are alert to every electoral event and through habit or normative commitment always participate. This group is not broadly representative of society but tend to be wealthier and more established socially, as well as more right of centre politically. But for the rest of the electorate, a variety of cues, prompts, or incentives will influence whether they participate or not. These involve both push and pull factors. Push factors include things like the effect of advertising and media coverage. Federal and provincial elections tend to have higher voter turnout because they get more media attention and generate more advertising. By contrast, municipal and stand alone referendums tend to produce much lower voter turnout because they do not generate as much media coverage or paid advertising. Pull factors include the role of political parties and civil society organizations in alerting voters to the electoral event and mobilizing them to participate. Again, these pull factors are stronger at the federal and provincial level than in municipal elections or referenda because party organization is not as present in the latter cases.

What all this means is that, compared to a federal or provincial general election, a stand alone referendum may struggle to gain sufficient media attention to drive public participation to comparable levels. A key factor will be the role that political parties play or do not play in the event. Research suggests that only political parties have the capacity to drive up voter turnout in the absence of sustained media attention, given their ability to cue voters about how to vote and mobilize them to participate in the voting opportunity. Referenda utilizing mail-in ballots also have some specific challenges in terms of voter turnout. For instance, as a rule, in-person electoral events tend to generate higher voter turnout than mail-in events, in part because the social pressure to vote is different and voters face more challenges in navigating the ballot in

4 mail-in events than when they attend an electoral event in person (where they can seek help if they need it). Without some attention to these issues from government in thinking through the participation challenges inherent in the kind of referenda they have committed to, the resulting voter turnout will undoubtedly be inequitable, skewed toward the more established segments of BC society and against the less privileged.

There a number of choices that the government could make to further equity in the voter participation in the referenda. First, it could encourage a strong role for political parties in mobilizing voters to participate, which could involve funding to parties to aid in such endeavors. Second, it should consult various American state governments like Oregon and California that have extensive experience with mail-in balloting for both general elections and referenda. Just of few of the American ‘best practices’ include postage paid return envelopes (which have a better return rate than ones without postage), the use of local constituency drop off centres for ballots (which may allow voters to get in-person help to understand the ballot, if they need it), and the organization of voting day ‘festivals’ in local constituencies to draw media and public attention to the referenda as an event.

Recommendations: that the government (a) structure a key role for political parties in mobilizing the electorate to participate and (b) explore American ‘best practices’ for aiding voter turnout when utilizing mail in ballots. c. Public education

The public needs to know that a referendum is taking place, what the substantive issue is that will be addressed with the referendum question, and the range of possible choices that can be made. The government’s approach to aiding the public in becoming informed so as to effectively participate in the referendum should be based on a realistic appraisal of what is both likely to occur and what is possible in the time frame available before the vote. It should also reflect a sober evaluation of the limits of past efforts around public education in recent voting system referenda in Canada.

The first question that needs to be answered is what does the public need to know to participate in this referendum? The previous voting system reform campaigns in BC that preceded the referendums in 2005 and 2009 appeared to assume that the public needed to understand the mechanics of the different voting systems in terms of how they were structured and how they produced an outcome. This was an uninformed and ultimately failed strategy. Why? Because publics across western countries do not understand their own voting systems, let alone any alternatives. For instance, despite using the single transferable vote form of proportional representation for over a century, people in Ireland generally cannot explain how votes are transferred to produce the outcome. People in Germany generally cannot explain how seats are allocated between single member and party list politicians for different political parties. Indeed, in Canada, the country that allegedly uses the ‘simple’ voting system, most people cannot explain how a party can turn 39% of the popular vote into a majority of seats. What people do generally understand is how to use their system to pursue the political

5 representation they desire. Thus people in Ireland know how rank choices on the ballot to aid their chosen political party or independent candidates. And people in Canada generally understand how to vote strategically for a party and avoid ‘wasting’ their vote. So public education should focus its themes on the outcomes generally produced by different systems and not the technical or mechanical details. As a rule, voters know very little about the details of any electoral institutions, not just the voting system. Instead, they count on the media and political parties to police the process and public concern tends to emerge only after such players have raised objections.

The recent spate of voting system referenda in Canada and the UK do offer some insights into what does not work in terms of informing the public. Here are three of the most important lessons. (1) Neutral campaigns from government election bodies simply stating that a referendum is taking place do not work. Evidence from recent experience in Canadian provinces suggests that such approaches left a majority of the public unaware that a referendum was going on, let alone what it was about. (2) Leaving the public education function to the media in the belief that they will offer both extensive and balanced coverage of the issue is not supported by past experience. Indeed, it did not come to pass in any of the provincial referenda. Instead, coverage was limited and biased toward maintaining the status quo. Indeed, the media in most cases acted as the unofficial ‘No’ campaign, with a clearly demarked political agenda of their own as concerned the issue. (3) As a rule, advocacy groups are not able to effectively reach the public in an educative capacity and can be uneven in the quality and reliability of the information they provide. In the recent Canadian voting system referendums, ‘No’ campaigns particularly typically offered up unsystematic research and cherry picked examples designed to confuse rather than inform voters in a balanced way. Survey research conducted during the campaigns reinforces the view that advocacy groups had little positive impact on public discussion in terms of what the public knew about the referendum.

With these insights into what the public needs to know and how previous campaigns failed, there are a number of choices the government can make to aid the process of public education for the referendum. First, it should recognize that most voters, indeed the great majority of voters, will not become informed about the details of voting systems. This does not mean that the public cannot or should not participate. What it means is that the government should adapt its approaches to the practical reality of how voters handle issue complexity. Some voters will want to know what is stake in the vote and this concerns how the choice of voting system may affect the broad workings of politics in the province – the potential outcomes, in other words. Issues like single party majority versus coalition government, more representative, competitive electoral processes versus processes that limit competition and representation, etc. But other voters will look to proxies for direction on how to vote, i.e. political parties or media or pundits they trust. As such, government should task an arm’s length, independent body (like Elections BC) to provide fact-based information to aid voters and structure a key role for political parties in helping voters come to grips with the issue.

Policy-wise this leads to a number of decisions. First, government should not provide public funds to advocacy groups for the referendum campaign. Past referenda have seen ‘No’

6 campaigns particularly simply funnel public money into negative advertising that often made dubious and ill-supported claims about how different voting systems work and the kind of results they would produce in BC, if adopted. Yes campaigns did offer more balanced, fact- based messages but did not appear to have much impact on levels of voter knowledge about the issue. Given the lack of demonstrated impact of these approaches, and the lack of quality in some of their efforts, it would not be good policy to spend public money in this way. Two, government should not assume that media will offer either substantive or balanced coverage of the issue. Instead, far from acting as a neutral arbiter of public discussion, past experience suggests that media operate with their own political objectives in mind. Third, it is imperative that the government place restrictions on campaign spending for the referendum to prevent distortions of the public debate. The potential for such distortions was clearly demonstrated by past experience in New Zealand, the UK, and the Canadian provinces. Finally, an arms-length public information body (like Elections BC) cannot simply be tasked with providing neutral messages stating that a referendum is occurring. Without some sense of what is at stake, voters may not understand why their participation is important. The public information needs to address the politics of the referendum choice (i.e. what is at stake).

Drawing from international experience, the government should empower Elections BC to prepare a public education campaign that is fact-based and balanced in its approach (New Zealand offers the best model of how to approach this). The government should also take up the practice utilized in California and Oregon where a body like Elections BC would produce a document that is mailed to every home containing the details of the referendum, the neutral information prepared by the the arm’s length public body, as well as submissions from any advocacy groups that wish to submit material. The American approach satisfies the fairness and equity concerns by assuring that all potential voters are exposed to a range of information about vote and why different groups think it is important.

Recommendations: that the government (a) empower Elections BC to prepare a fact-based education campaign, (b) put in place restrictions on third party spending for the referendum campaign, (c) structure a key role for political parties in the education campaign, with funding to support their efforts, (d) not fund advocacy groups, and (e) produce a document that can be mailed to every home in BC containing information about the referendum and include statements from any and all advocacy groups that wish to submit material.

2. Design and Implementation issues

In the event that a majority of voters decide to adopt a new voting system for British Columbia the government will need to make a number of design and implementation decisions, specifically around questions of districting and the total number of MLAs to be elected under the new rules. For instance, concerns have been raised in the media about the potential impact of any new system on the level of non-urban representation in the province or the size of constituencies both in terms of population and geography. As such, the government will need a plan to address any concerns that arise from designing and implementing the new system. At this point, the government should offer up ‘decisions without details’. In other words, it should

7 spell out the general process by which various issues will be addressed and resolved without necessarily offering solutions to specific issues at this point. This is important because public input will be necessary after the referendum to assess just how to manage the trade-offs that will be required to set up any new system. a. Balancing rural versus urban representation

If a proportional voting system is chosen as the alternative to BC’s traditional single member first past the post system then new constituencies will have to be designed. Regardless of the specific proportional system chosen, such new constituencies will have to be larger than the present single member ones, in terms of both population and geography. Anticipating such a process, some in the media have argued that the resulting redistricting process would see a reconfiguration of seats that would reduce non-urban representation. This confuses a number of issues. Setting aside the referendum issue, BC’s existing electoral map already departs from the Canadian Supreme Court’s guidelines in terms of the acceptable level of deviation from voter parity in the design of less populated constituencies. As such, this map could be challenged legally and the government could be forced to bring it into line with the court’s rules, or the government could decide as a matter of principle that BC’s map should adhere to the letter of law and direct changes to be made without legal prompting. All this is separate from the issue about whether BC should change its voting system. There is nothing intrinsic to adopting a new voting system that would or must necessarily change the ratio of urban to rural seats.

Bracketing the court-related aspects of the rural/urban division for a moment, the government could simply signal that it intends to maintain the existing proportionate division between rural and urban representation in the design of the new electoral map. In other words, if the number of seats to be used in the new system remains the same as now, the government could state that there will be no change in the allocation between regions, though, of course, the design of the constituencies within regions would have to change. And if public feedback after the referendum suggested that an increase in the number of representatives is preferred to help make the change, here too the government could assure voters that the proportions of seats would stay the same, even as the number of seats increased. b. Designing new constituencies

As noted, a new voting system will require a new electoral map and means of deciding how to design it. In the interests of historical continuity and increasing the legitimacy of the process, the government should utilize the province’s traditional electoral boundary commission approach to gain public input and make decisions about where to draw the lines for the new constituencies. The commission will need clear instructions from the government on a number of issues (e.g. to maintain the existing rural/urban proportionate breakdown of seats, or not; establishing the breakdown of single member versus multi-member seats, or the mix of multi- member seats; etc.) but within its mandate should be free to take public direction and make decisions for the government to act on. The commission can take the measure of public opinion

8 around whether to increase the number of MLAs to better aid the transition to the new system, or not, as well as which regions should be combined together to create multi-member constituencies and/or larger single member ridings, depending on the system chosen. The commission will need a strict timeline and the necessary resources to be able to complete its work in time for Elections BC to prepare for a general elections using the new map and voting system.

Recommendations: that the government (a) signal its position about the balance of urban/rural representation and (b) commit to using the province’s traditional electoral boundary commission approach to designing the new voting system.

Conclusion

The recommendations above are designed to enhance the fairness, equity and legitimacy of the referendum process and possible design/implementation of a new voting system for BC. But it must be noted that such values are themselves the terrain of political contestation. In other words, as has become clear already in the political and media responses to nearly all government decisions on this issue thus far, contesting the fairness and legitimacy of the process is likely going to be a key political strategy at work in the referendum, particularly from those opposed to change. So gaining public approval from everyone for decisions surrounding the vote and implementing the new system is clearly not likely. Still, setting aside purely partisan-driven opposition, there is considerable room to appeal to the general public’s sense of good will and fair play. And the above recommendations will go some distance to creating a more fair referendum process, both substantively and at the level of appearances.

9 Biography

Dennis Pilon was born in Vancouver and has lived all over British Columbia, including the lower mainland (Vancouver, Burnaby, New Westminster, Coquitlam, and Surrey), Vancouver Island (Port McNeil, Victoria), the central interior (Williams Lake), the northern interior (Smithers) and the far north (Cassiar). He has a BA in Sociology and History and an MA in History from Simon Fraser University and PhD in Political Science from York University. He has written extensively about the political history of British Columbia, co-editing the collection British Columbia Politics and Government with Michael Howlett and Tracy Summerville, as well as the province’s many instances of democratic reform. These include campaigns for municipal voting system reform around World War I, in the 1930s, and from the 1970s on, particularly in the city of Vancouver, as well as provincial reform efforts in the 1950s, 1970s, and from the 1990s on. He is also an expert on voting systems and their use, as well the process of voting system reform across western industrialized countries.

Relevant publications

Books

Wrestling with Democracy: Voting Systems as Politics in the Twentieth Century West, (Toronto: University of Toronto Press, 2013), 394pp.

Politics and Government in British Columbia, (Toronto: Emond Montgomery, 2009), co-edited with Michael Howlett and Tracy Summerville, 404pp.

The Politics of Voting: Reforming Canada’s Electoral System, (Toronto: Emond Montgomery, 2007), 209pp.

Academic journal articles

“Canada's Voting System in Comparative Perspective,” Journal of Parliamentary and Political Law, 8, (July 2015), 53-64.

“Electoral Reform: Here’s Your Evidence Mr. Trudeau,” Inroads, 37 (Summer/Fall 2015), 51-60.

“The 2005 and 2009 Referenda on Voting System Change in British Columbia,” Canadian Political Science Review, 4: 2-3 (June-September 2010), 73-89.

“Investigating Media as a Deliberative Space: Newspaper Opinions about Voting Systems in the 2007 Ontario Provincial Referendum,” Canadian Political Science Review, 3: 3 (September 2009): 1-23.

“Explaining Voting System Reform in Canada: 1874 to 1960,” Journal of Canadian Studies, 40: 3 (Autumn 2006), 135-61.

10

Book chapters

“Assessing Gordon Campbell’s Uneven Democratic Legacy in British Columbia,” in Tracy Summerville and Jason Lacharite (eds.), The Campbell Revolution? Power, Politics and Policy in British Columbia from 2001 to 2011, (Kingston-Montreal: McGill-Queen's University Press, 2017), 37-60.

“Voting Systems and Party Politics,” In Alain-G. Gagnon and A. Brian Tanguay (eds.), Canadian Parties in Transition: Recent Trends and New Paths for Research, 4th edition, (Toronto: University of Toronto Press, 2017), 217-49.

“British Columbia: Right-wing Coalition Politics and Neoliberalism,” in Bryan M. Evans and Charles W. Smith (eds.), Transforming Provincial Politics: The Political Economy of Canada's Provinces and Territories in the Neoliberal Era, (Toronto: University of Toronto Press, 2015), 284-312.

“Democracy, BC-style,” in Michael Howlett, Dennis Pilon and Tracy Summerville (eds.), Politics and Government in British Columbia, (Toronto: Emond Montgomery, 2010), 87-108.

“Understanding Electoral Systems,” in Heather McIvor (ed.), Election, (Toronto: Emond Montgomery, 2009), 35-56.

“Elections and Electoral Reform,” in Joan Grace and Byron Sheldrick (eds.), Canadian Politics: Democracy and Dissent, (Toronto: Pearson, 2006), 229-50.

“The Uncertain Path of Democratic Renewal in Ontario,” in Henry Milner (ed.), Steps Toward Making Every Vote Count: Electoral System Reform in Canada and its Provinces, (Peterborough: Broadview, 2004), 249-66.

“The History of Voting System Reform in Canada,” in Henry Milner (ed.), Making Every Vote Count: Reassessing Canada’s Electoral System, (Peterborough: Broadview, 1999), 111-21.

Op/eds

“Change the Voting System Without a Referendum,” Vancouver Sun, May 29, 2017. “Our Entire Electoral Reform Debate is Based on a Fear of Voters,” iPolitics, Dec 12, 2016. “The Dogfight over Electoral Reform is about Power - Full Stop,” iPolitics, Nov 17, 2016 “You Can't Hide Behind the Constitution to Spare Us Electoral Reform,” National Post, Feb 1, 2016.

11